Absolutely on the money!

The problem with retiring redundant, monolithic applications with tools which improve agility and efficiency is that all too often you need a multiplicity of tools e.g. CRM, analytics, rules engines, process and situational needs-based tools. These almost always come from different vendors or companies they have acquired and bolted on. The result is that you replace one form of complexity with another.

Only by deploying one tool which does everything, can the objectives Kristin outlines be achieved: An agile and efficient environment.

I agree that security, scalability, performance, complexity, regulatory compliance and integration should simply be managed by choosing the right tool. This puts the control back in the hands of the business, who are then in a much better position to respond rapidly to customer needs.

Banks should be building applications for change - otherwise they are simply replicating the legacy that they are attempting to escape.

Gartner are spot on with this! However for financial services firms to engage they must first apreciate their problem and understand what they need to do to address it (A plan is good) and i think some imagination to utilise the available technology. As a guide think customer services and do not deviate from that thought

As bank CIOs attempt to open up access to their systems by providing clients with APIs for their data and transaction processing systems, they will find that the power they can offer through APIs will be restricted, both in functionality and client access. This is because transactional facilities by their nature come loaded with risk management and regulatory requirements that financial institutions cannot simply offload on to clients.

 The provision of API access to raw transactional systems is just not feasible.  One solution is to create an application layer between the processing systems and the public APIs to provide business logic – typically in the form of purposely-engineered payment programs – specifically designed to offer valuable services that can be effectively overseen and controlled from a risk and compliance perspective.

I agree with Gartner's view that banks can better serve their customers by following the approach so well illustrated by the example of the automatic mortgage refinancing app. I can think of many other examples viz. rejigging fixed deposit portfolios to maximize yield whenever interest rates undergo changes in nations like India where corporates and individuals alike keep a lot of money in fixed deposits that carry as high as 10% p.a interest rates while carrying zero risk. I personally use an Excel spreadsheet to simulate whether it's worth breaking an FD earning (say) 8% and reinvesting it when the rate has gone up to (say) 9.5% even at the cost of losing 1% point interest for premature withdrawal OR leave it where it is even if it earns 1.5% point lower interest. If my bank introduced a smartphone / PC web app to keep doing this in the background and gives me suggestions proactively, I'd use this app to my great benefit. 

However, as I'd pointed out in this post, banks might actually lose money by giving me such an app. More than "loss of control" or any other factor listed by Gartner, I surmise that it's this fear of loss of revenue and / or profit that acts as a greater barrier in banks' adoption of the approach recommended by Gartner.